Evaluating the spatial relationship of event‐related potential and functional MRI sources in the primary visual cortex

Abstract

The integration of electroencephalogram (EEG) recordings and functional magnetic resonance imaging (fMRI) can provide considerable insight into brain functionality. However, the direct relationship between neural and hemodynamic activity is still poorly understood. Of particular interest is the spatial correspondence between event-related potential (ERP) and fMRI sources. In the current study we localized sources generated by a checkerboard stimulus presented to eight subjects using both EEG and fMRI. The location of the sources of the visual evoked potential (VEP) were estimated at each timepoint and compared to the location of peak fMRI activity. In the majority of participants we found that the N75 dipole location coincides with a region of positive blood oxygenation level-dependent (BOLD) activation and the P100 dipole location coincides with a region of negative BOLD activation. These findings demonstrate the importance of including the negative BOLD response in combined EEG/fMRI studies.

Voxel-wise information theoretic EEG-fMRI feature integration

“…S5) indicate that, at the level of the false discovery rate threshold used, there were no brain regions in which the P100 amplitude was informative about the HRF amplitude. This is in accordance with our previous work, where the absolute values of activity dependence were low and not significantly different to zero , see also (Whittingstall et al, 2007). This null result highlights the need for a more detailed investigation of the EEG signal space to identify features which carry information about the haemodynamic response.…”

Relationship of the BOLD Signal with VEP for Ultrashort Duration Visual Stimuli (0.1 to 5 ms) in Humans

“…In addition, fMRI activation maps show that in response to brief light stimuli, most of the activated voxels are located in the primary visual cortex (Figure 1). This is in agreement with a similar study where it was reported that the early VEP component is localized closer to the V1 BOLD response than the later VEP component (Whittingstall et al, 2007(Whittingstall et al, , 2008. On the other hand, other studies have shown that different evoked potential deflections correlate best with the BOLD response, for example, P3a is related to BOLD signal changes in the right fusiform and left superior temporal gyrus for a facial emotion recognition task (Johnston et al, 2005); N170 for face and house visual stimuli have been found to correlate well with hemodynamic responses in various brain areas in the temporaloccipital lobes (Iidaka et al, 2006); and N400 and BOLD signals were linked in the superior temporal gyrus for a semantic priming task (Matsumoto et al, 2005).…”

Peripherally Presented Emotional Scenes: A Spatiotemporal Analysis of Early ERP Responses

“…Indeed, visual information coming from the right visual hemifield is conveyed to the left occipital visual areas (and vice-versa for the left visual hemifield) via both ipsilateral and contralateral pathways. Such predominant contralateral activity in response to peripherally presented pictures was already found with fMRI [46], magnetoencephalography [47] or visual evoked potentials [48] but never at such eccentric locations. The lateralization of the early visual activity could also be related to a more extended and direct innervation of secondary and more lateral areas by the magnocellular pathway, as suggested by a few studies in animals ( [49] in monkeys; [50] in cats).…”